Zobrazit v EDS

VANTAGE: An open‐source application for visualisation, navigation and analysis of video and ancillary time‐series data in bio‐logging studies

Uloženo v:
Podrobná bibliografie
Název: VANTAGE: An open‐source application for visualisation, navigation and analysis of video and ancillary time‐series data in bio‐logging studies
Autoři: Stefan Schoombie, Tagen de Wet, Stefan Britz, Andrew D. Lowther, W. Chris Oosthuizen
Zdroj: Methods in Ecology and Evolution, Vol 16, Iss 5, Pp 912-920 (2025)
Informace o vydavateli: Wiley, 2025.
Rok vydání: 2025
Témata: accelerometer, machine learning, behavioural analysis, Ecology, Evolution, penguin, QH359-425, animal‐borne video, OpenCV, QH540-549.5
Popis: Animal‐attached devices, or bio‐loggers, that record data on multiple channels are frequently used to study the movement of free‐ranging animals. In recent years, the deployment of animal‐borne video cameras in addition to other time‐series loggers, such as accelerometers and depth sensors, has become popular. Visual observations from animal‐borne cameras provide ideal ground‐truth data when behavioural inferences are made from concurrently deployed bio‐loggers. However, such deployments generate large amounts of data that are often not synchronised and may be difficult to analyse because open‐source software tools facilitating simultaneous visualisation of multiple data channels along with video streams are lacking. Here we present VANTAGE (Video Analysis and Navigation Tool for Advanced Graphical Exploration, available at https://github.com/sschoombie/VANTAGE), a new open‐source graphical user interface written in Python. VANTAGE aims to simplify time‐consuming data processing and analyses through (1) concurrent visualisation of video footage and accompanying bio‐logger time‐series data, (2) temporal synchronisation of the different data streams, (3) efficient annotation across data types, and (4) video enhancement and machine learning functionalities. We illustrate both manual and automated synchronisation of animal‐borne video and accelerometer‐depth data (obtained from foraging penguins) using VANTAGE. Once synchronised, VANTAGE allows for quick and easy navigation through the data to areas of interest, with behavioural observations labelled on one data type reflecting in all the data series. The labelled time‐series data can then be exported for further analyses (e.g., in machine learning applications). VANTAGE allows the integration of video and other time‐series data recorded at the same time, allowing more efficient analysis and labelling of the data. Bio‐logging studies often use data from multiple loggers, and analysing these data can be challenging when not properly synchronised. VANTAGE aims to simplify this synchronisation process through visualisation of multiple data streams.
Druh dokumentu: Article
Jazyk: English
ISSN: 2041-210X
DOI: 10.1111/2041-210x.70026
Přístupová URL adresa: https://doaj.org/article/85170f856d034a068320accc7c50393e
Rights: CC BY NC
Přístupové číslo: edsair.doi.dedup.....55dfaa2966bd8da7a926e1696053ff7d
Databáze: OpenAIRE
Popis
Abstrakt:Animal‐attached devices, or bio‐loggers, that record data on multiple channels are frequently used to study the movement of free‐ranging animals. In recent years, the deployment of animal‐borne video cameras in addition to other time‐series loggers, such as accelerometers and depth sensors, has become popular. Visual observations from animal‐borne cameras provide ideal ground‐truth data when behavioural inferences are made from concurrently deployed bio‐loggers. However, such deployments generate large amounts of data that are often not synchronised and may be difficult to analyse because open‐source software tools facilitating simultaneous visualisation of multiple data channels along with video streams are lacking. Here we present VANTAGE (Video Analysis and Navigation Tool for Advanced Graphical Exploration, available at https://github.com/sschoombie/VANTAGE), a new open‐source graphical user interface written in Python. VANTAGE aims to simplify time‐consuming data processing and analyses through (1) concurrent visualisation of video footage and accompanying bio‐logger time‐series data, (2) temporal synchronisation of the different data streams, (3) efficient annotation across data types, and (4) video enhancement and machine learning functionalities. We illustrate both manual and automated synchronisation of animal‐borne video and accelerometer‐depth data (obtained from foraging penguins) using VANTAGE. Once synchronised, VANTAGE allows for quick and easy navigation through the data to areas of interest, with behavioural observations labelled on one data type reflecting in all the data series. The labelled time‐series data can then be exported for further analyses (e.g., in machine learning applications). VANTAGE allows the integration of video and other time‐series data recorded at the same time, allowing more efficient analysis and labelling of the data. Bio‐logging studies often use data from multiple loggers, and analysing these data can be challenging when not properly synchronised. VANTAGE aims to simplify this synchronisation process through visualisation of multiple data streams.
ISSN:2041210X
DOI:10.1111/2041-210x.70026